The AI Execution Pipeline allows for the seamless integration of multiple AI modules to enhance the capabilities of your chatbots. By combining different modules, you can create dynamic workflows that adapt to various user interactions. When a query comes in from user, it passes to the first module of the pipeline and will be fall back to the next one if AI is not able to handle that specific query.
Available AI Modules for Execution Pipeline:
Classification Model
LLM (Large Language Model)
Cognitive Search
Agent M
Classification Model
This module follows the traditional intent + entities model for understanding user input. It helps in categorizing user queries into predefined intents and extracting relevant entities.
LLM
The LLM module uses enhanced natural language understanding skills to deliver more complex responses. It can handle complex language patterns and context to generate human-like interactions.
Cognitive Search
With the help of this module, the chatbot can respond to user inquiries by retrieving data from papers, PDFs, webpages, and other sources. It provides thorough search features to retrieve relevant and reliable data.
Agent M
Agent M empowers users to create custom LLM agents tailored for specific tasks. It allows for the development of specialized agents with focused capabilities, enhancing the chatbot’s performance in particular domains.
